Volume 9, Issue 3, Pages 223-234 (March 2011) HIV-1 Utilizes the CXCR4 Chemokine Receptor to Infect Multipotent Hematopoietic Stem and Progenitor Cells  Christoph C. Carter, Lucy A. McNamara, Adewunmi Onafuwa-Nuga, Mark Shackleton, James Riddell, Dale Bixby, Michael R. Savona, Sean J. Morrison, Kathleen L. Collins  Cell Host & Microbe  Volume 9, Issue 3, Pages 223-234 (March 2011) DOI: 10.1016/j.chom.2011.02.005 Copyright © 2011 Elsevier Inc. Terms and Conditions

Figure 1 Total Number of Bone Marrow Mononuclear Cells Is Correlated with Year of Diagnosis in HIV+ Donors with High Viral Loads (A) Complete blood count of HIV+ donors with high viral loads (1–6 and A1–A3, 61,000–202,000 copies/ml) and low viral loads (7–16, <48 copies/ml). (B) The number of purified bone marrow cells versus year of diagnosis for high-viral-load donors. (C) CD4 counts of high-viral-load donors at the time of bone marrow aspiration plotted as a function of year of diagnosis. (D) As (B), but with an outlier with a high mononuclear cell yield (Donor 5) removed. (E) The number of purified bone marrow cells versus year of diagnosis for low-viral-load donors. (F) CD4 counts of low-viral-load donors at the time of bone marrow aspiration plotted as a function of year of diagnosis. Cell Host & Microbe 2011 9, 223-234DOI: (10.1016/j.chom.2011.02.005) Copyright © 2011 Elsevier Inc. Terms and Conditions

Figure 2 CXCR4-Tropic HIV Envs Infect CD133+, CD34+ HSPCs (A) Schematic of HIV-7SF-GFP construct and generic HIV envelope plasmid used to construct viruses used in (B)– (D). (B and C) Flow cytometric analysis of cord-blood-derived CD133+ cells infected with a minimal HIV (HIV-7SF-GFP [Yam et al., 2002]) that lacked expression of HIV gene products and was pseudotyped with the indicated HIV Env. In this construct, GFP is expressed from a heterologous promoter (Yam et al., 2002). Cells were analyzed 3 days postinfection. The left panels show the GFP+ gating. These events were overlaid in green on the right plots, which also show staining of the total cell population in gray. Isotype control staining is shown in the top panel. The percentage of GFP+ cells that were CD34HighCD133+ is shown in the upper right-hand corner. (D) Summary plot of HIV-1 Env data. Results are compiled from ten experiments with CCR5-tropic Env proteins (94UG114.1.6, eight replicates; YU-2, one experiment; ZM53M.PB12, one experiment) and 16 experiments with CXCR4 and dual-tropic Env proteins (HXB2, eight replicates; 89.6, seven replicates; 92HT593, one experiment); error bars are standard error of the mean. ∗∗∗p < 0.0001. (E) Upper panel, schematic of full-length HIV used in lower panel. Lower panel, flow cytometric analysis of cord-blood-derived CD133+ cells 3 days postinfection with full-length, wild-type HIVs. NL4-3 has an X4-tropic Env and 94UG114.1.6 has an R5-tropic Env. Gag+ cells were gated in the left panels, and these events (red dots) were overlaid on CD34 versus CD133 plots in the right panels to determine the immunophenotype of the infected cells compared with the total cell population (gray dots). Isotype control staining is shown in the top panel. The percentage of Gag+ cells that were CD34+CD133+ is shown in the upper right-hand corner. See also Figure S1. Cell Host & Microbe 2011 9, 223-234DOI: (10.1016/j.chom.2011.02.005) Copyright © 2011 Elsevier Inc. Terms and Conditions

Figure 3 CXCR4-Tropic HIV Envs Infect HSPCs with the Capacity to Form Multilineage Colonies (A) Schematic of HIV-7SF-GFP construct and generic HIV envelope plasmid used to construct viruses used in (B)– (E). (B) Flow cytometric analysis of cord-blood-derived CD133+ cells infected with HIV-7SF-GFP pseudotyped with HXB (X4-tropic) or YU2 (R5-tropic) Env proteins and purified by flow sorting. (C) Example colonies identified after culturing cells isolated as shown in (B) for 14 days. Phase contrast and epifluorescence microscopy are shown (erythroid, CFU-E; myeloid, CFU-GM; or multilineage, CFU-GEMM). (D) Quantification of colony formation for the experiment shown in (C) and for two similar experiments using other HIV Envs, as indicated (erythroid, CFU-E; myeloid, CFU-GM; or multilineage, CFU-GEMM). The total number of colonies from six replicate wells is displayed. (E) Summary of colony formation results. Data were compiled from five independent experiments using six different Env proteins. The average normalized number of colonies observed with CXCR4- or dual-tropic Env versus CCR5-tropic Env is depicted. Error bars represent standard deviation and p values were determined using the two-tailed Student's t test (erythroid, CFU-E; myeloid, CFU-GM; or multilineage, CFU-GEMM). (F) Summary table of the ability of HIV-1 Envs of different tropism to infect CD133+CD34High and multipotent HSPCs. Cell Host & Microbe 2011 9, 223-234DOI: (10.1016/j.chom.2011.02.005) Copyright © 2011 Elsevier Inc. Terms and Conditions

Figure 4 Infection of CD34HighCD133+ HSPCs Is Dependent on CD4 and CXCR4 (A) Flow cytometric analysis of cord-blood-derived CD133+ HSPCs or CEM T cells preincubated with anti-CD4 antibody (L3T4, 20 μg/ml) or control antibody and then infected with HIV-7SF-GFP pseudotyped with 89.6 Env. The cells were analyzed 3 days postinfection. The numbers in the right panels are fold inhibition of infection. (B) Flow cytometric analysis of cord-blood-derived CD133+ HSPCs preincubated with control antibody or antibody to CD4 (L3T4, 20 μg/ml) and then infected with HIV-7SF-GFP pseudotyped with 89.6 Env. The cells were analyzed 3 days after infection. GFP+ (red) cells are overlaid on the total population (gray). (C) Flow cytometric analysis of CEM-SS (CXCR4-expressing) or CEM-R5 (CXCR4- and CCR5-expressing) cells treated with 10 μg/ml AMD3100 or an equal volume of DMSO and then infected with HIV-7SF-GFP pseudotyped with the dual-tropic HIV Env 92HT593. The cells were analyzed 3 days postinfection. The percent GFP+ cells are shown in the numbers above each gate. Results are representative of two independent experiments. (D–F) Flow cytometric analysis of cord-blood-derived CD133+ cells infected with HIV-7SF-GFP pseudotyped with the dual-tropic 89.6 (D) or 92HT (F) HIV Envs or the R5-tropic YU-2 HIV env (E) in the presence and absence of 20 μM maraviroc (R5-blocking) or 10 μg/ml AMD3100 (X4-blocking). The cells were analyzed 3 days after infection. GFP+ cells (red) are overlaid on the total population (gray). (G) Flow cytometric analysis of CD133+ UCB cells expanded 4–7 days in STIF media and then stained for the indicated cell surface markers. Samples stained with isotype control antibodies for CD133 and CD34 (top) or with antibodies to CD133 and CD34 but isotype control antibodies for CD4, CXCR4, and CCR5 (middle) were included to determine gating. Numbers indicate the percent of cells falling within each gate. (H) Summary graph showing percent of total CD133+CD34High cells that express both of the indicated receptors for three independent experiments; mean and standard deviation (error bars) are shown. ∗p < 0.03, paired t test. See also Figure S2. Cell Host & Microbe 2011 9, 223-234DOI: (10.1016/j.chom.2011.02.005) Copyright © 2011 Elsevier Inc. Terms and Conditions

Figure 5 Infection of Multipotent Cells Requires CXCR4 (A and B) Colony formation by cord-blood-derived HSPCs treated with 20 μM maraviroc (A), 10 μg/ml AMD3100 (B), or an equal volume of DMSO, infected with HIV-7SF-GFP pseudotyped with the dual-tropic HIV Env 89.6 and purified by FACS. Two blinded counters analyzed the colonies 14–18 days after plating in methylcellulose medium. Data are represented as number of colonies per 100 cells plated. The mean of the two scorers' counts is shown. (C and D) Colony formation by CD133+ UCB mock-infected in the presence of 20 μM maraviroc (C) or 10 μg/ml AMD3100 (D) or an equal volume of DMSO and plated in methylcellulose 3 days after exposure. Colonies were scored after 2 weeks by two blinded counters. The mean of the two scorers' counts is shown. (Erythroid, CFU-E; myeloid, CFU-GM; or multilineage, CFU-GEMM.) Cell Host & Microbe 2011 9, 223-234DOI: (10.1016/j.chom.2011.02.005) Copyright © 2011 Elsevier Inc. Terms and Conditions

Figure 6 CXCR4-Tropic HIV Can Infect HSCs that Stably Engraft and Generate Multiple Lineages in NOD/SCID IL-2Γnull Mice (A) Upper panel, schematic of HIV-7SF-GFP and HXB envelope plasmid used to construct virus used in lower panel. Lower panel, flow cytometric analysis of cord-blood-derived CD133+ HPCs infected with HIV-7SF-GFP pseudotyped with HXB Env after 3 days in culture. Cells in the right panel were sorted for GFP positivity. (B) Flow cytometric analysis of peripheral blood from transplanted mice 16–18 weeks posttransplant. Leukocytes that were positive for human CD45 and negative for mouse CD45 were gated in the left panels. The right panels show staining of the indicated markers within this subpopulation. (CD33, myeloid; CD3, T cell; CD19, B cell.) (C) Flow cytometric analysis of peripheral blood cells harvested 20 weeks posttransplantation in animal 9. (D) Flow cytometric analysis of bone marrow and spleen from animal 6 after 26 weeks. (C: untreated control animal.) See also Figure S3. Cell Host & Microbe 2011 9, 223-234DOI: (10.1016/j.chom.2011.02.005) Copyright © 2011 Elsevier Inc. Terms and Conditions